A thread is a unit of flow executed in a process that processes information necessary to handle a request for a particular service. The thread processes the particular service in response to the request and shares information of the processed particular service with other threads.
A service, i.e., a caller service, processed by a thread may call another service, i.e., a callee service, to be processed by another thread, and there are two existing models: synchronous and asynchronous models.
According to the synchronous model, since the caller service must wait until the callee service is finished, the thread processing the caller service is at a suspending state until a processing of the callee service is finished, resulting in an inefficiency of the processing of the caller service.
Meanwhile, according to the asynchronous model, the thread may process an unprocessed part of the caller service while the callee service is being processed. However, as shown in FIG. 1, the thread cannot process other services, resulting in an inefficiency of the management of the services.
Accordingly, the present invention proposes a method and a computing device for increasing throughputs of services to be processed by threads.